I have a pretty good idea of what I could end up doing for my thesis project. I want to amplify the bodily expression of the viewers by capturing their movement and projecting an enhanced audiovisual version of it. For this, I will use Machine Learning to recognize expressive movements (not only about emotion of a pose, also analyzing extremities by themselves and taking into account the speed of the movement) and later create visuals and sounds that go accordingly.
Ideas out there
So far, I have not found a project that does what I want to achieve; but there are some that do approach certain aspects of my goal.
- Chris Milk – Treachery of Sanctuary
> I’m really interested in trying to figure out ways that technology… as technology evolves, it can be integrated into human experiences, and can you use technologies to produce real human emotions in people.
This 3 panel installation (Birth / Death / Transcendence) uses Kinect to track the body of the audience and morph them to convey certain ideas and generate emotions
- Kat Sullivan – Parsing our Silent Language
Kat’s thesis used Machine Learning on body language to predict the emotion of the user. It is a great starting point for the ML aspect, but for my goals, I need to take different aspects into account. - Brandon Newberg did a related project last semester (ML algorithm trained on people dancing or not) (need to talk to him)
-
Wayne McGregor – Choreographic Language Agent : fascinating way of portraying choreography over time
Research : Dance and Choreography
I am dividing my research in several parts. This week I did most of my research on dance and choreography. While a lot of these works are about algorithmically generated instructions for performers (and I want to do something that’s more like the opposite), it is useful to know what’s been done, the methods used and see the actual results.
- Labanotation : how to write movement?
it’s a lot about form, clarity of form. it doesn’t come from looking at yourself but really sensing where you are and the distance of your body parts from each other, and from your colleagues and from the universe
- Merce Cunningham (wiki)
> In his performances, he often used the I Ching in order to determine the sequence of his dances
algorithmically generated dance! > algorithmic choreography
DANCE: collaboration with Phillip Glass and Sol Lewitt (1979) Walker Art Center – A Look…, The Guardian – Review, Whitney exhibition
- Kate Sicchio : AMAZING works!
- Hacking the Body
- Sound Choreographer Body Code
- Hacking Choreography 2.0
- Hacking Choreography
- Eyeo 2017 presentation
Other Inspirations
- Inori – Prayer : high accuracy facial projection mapping. I would love to be able to project on top of the audience
- Crypt of the NecroDancer : fun rhythmic twist on a game
- Posemaniacs : 30 second challenges to learn how to draw human shapes. it taught me a lot and helps to distill a pose into its basic flow
- Ballet Rotoscope : beautiful reinterpretations of movement